How To Map The Sources Shaping AI Answers In Your Industry

Marketers monitoring brand visibility in AI-powered search results may be missing a critical layer of data: the specific sources that AI engines cite when generating answers. That was the central argument Gintare Rimolaityte, Chief Commercial Officer at Trendos, made during a Search Engine Journal webinar published on 21 September 2026.

Rimolaityte's presentation centred on the distinction between appearing in traditional search results and being cited as a source within an AI-generated answer. According to the webinar summary, her core point "was not to abandon SEO" but rather "to stop treating search rankings as a complete picture of AI visibility."

Citation Patterns Vary by Industry

Drawing on summer 2026 data from Trendos' analysis, Rimolaityte examined citation patterns across three industry verticals: consumer products and retail, IT and solution services, and communication services. Community and user-generated sources were described as prominent across all three categories, though the remaining mix varied by sector.

For IT and solution services, review sites, editorial sources, and community content featured prominently, with examples including G2, Reddit, YouTube, and technology publications. For communication services, community content and independent editorial sources were highlighted as important. In consumer products and retail, brand and retail-owned sources played a substantial role alongside communities.

The webinar summary notes these were "observed citation patterns in Trendos' analysis, not a formula that guarantees inclusion in an AI answer."

Engine-by-Engine Differences

Rimolaityte cautioned against drawing broad conclusions from a single engine's behaviour. She used Reddit as a specific example, noting that reduced prominence in ChatGPT did not mean it had disappeared from other engines examined. The source summary also notes she described differences in the balance of community, editorial, and brand-owned sources across Google AI Overviews, Perplexity, ChatGPT, and Gemini, with Gemini's source mix described as particularly dependent on industry.

Turning Competitor Citations Into a Worklist

Rimolaityte proposed a structured audit process beginning with 10 to 20 questions customers might ask while choosing a product or provider. Marketers are advised to run those questions in relevant engines, record the cited sources, and then examine pages where competitors appear and their own brand does not.

She explained why those gaps matter directly: "If your competitor appears in those sources and you don't, this is your opportunity to get the mention on the source because this source is already being used to get the answer, right?"

She warned against treating every discovered domain equally, recommending that marketers examine citation share and which source categories drive answers for their specific industry, rather than targeting large numbers of editorial sites indiscriminately.

Video, Transcripts, and Community Participation

During the Q&A portion of the webinar, Rimolaityte addressed questions about video content, saying brand and founder content could also be useful in AI citations. She emphasised concise question-and-answer formats and accurate transcripts, and recommended tracking whether a channel or its content appears among the sources for the prompts being monitored.

On community participation, she recommended honest participation from a company or employee voice rather than disguised accounts, and noted that community discussions can hold value beyond their potential to generate AI citations.

Audit Framework for Teams and Agencies

The webinar outlined a repeatable process: choose a manageable set of buying questions, record answers and cited sources by engine with consistent metadata, separate source types, inspect competitor gaps for legitimate next steps, and assign owners before repeating the checks to track changes over time. An agency attendee who asked about implementing recommendations without direct access to a client's properties was advised to prepare a concise, client-specific report identifying gaps and providing a to-do list.